Long‐term regular follow‐up yields important benefits in forensic psychiatric patients with psychosis: A 2‐year longitudinal study on psychopathology, general health, quality of life, and functioning

Abstract

Forensic psychiatric patients often present with severe psychopathology, functional impairment, and poor quality of life. Longitudinal follow‐up studies in forensic psychiatric outpatient populations remain limited. The objective is to assess psychopathology, quality of life, global functioning, and general health in forensic psychiatric outpatients compared with psychiatric patients without a forensic history over 24 months. A prospective longitudinal study was conducted with 120 participants (63 forensic and 57 psychiatric outpatients). Assessments were conducted at baseline, 6, 12, and 24 months using validate psychometric tools, PANSS, GHQ‐28, WHOQOL‐BREF, and GAF. Longitudinal changes within and between groups were analyzed. Most participants were male (82.5%), single (64.2%), unemployed (72.3%), and living in urban areas (93.3%). Forensic psychiatric outpatients reported less social support than psychiatric patients (68.3% vs. 89.5%). Most had a history of incarceration (93.8%). Schizophrenia was the predominant diagnosis (98.3%). Psychopathology, functioning, quality of life, and general health improved significantly over time in both groups (p < 0.001). Positive symptoms improved more in psychiatric patients (mean change −8.96). Psychiatric patients scored significantly higher in the WHOQOL‐BREF “environment” domain at baseline and at 6 months. Greater improvement in anxiety/insomnia symptoms was observed in psychiatric patients (mean change −6.44). Long‐term outpatient psychiatric follow‐up was associated with significant improvements in symptom severity, functioning, quality of life, and general health in both groups. However, forensic psychiatric outpatients continued to experience greater disadvantages, highlighting the need for tailored, long‐term, community‐based interventions in forensic mental health care.

Highlights.

• A 24‐month longitudinal study comparing forensic and general psychiatric outpatients.

• Long‐term outpatient follow‐up improves outcomes in forensic psychiatric patients.

• Psychopathology, functioning, and quality of life improved for both groups over time.

• Forensic psychiatric patients showed persistent social and clinical disadvantages.

• Findings support long‐term, community‐based forensic mental health care and support.

1. INTRODUCTION

Forensic psychiatry concerns the intersection between mental health and the criminal justice system, addressing clinical evaluation, treatment, and the risk assessment and management of individuals whose psychiatric disorders contribute to criminal behavior [1, 2]. In these cases, medical and legal oversight is required. Legal decisions may involve mandatory psychiatric evaluations, continuing treatment orders, or security regulations, which distinguish forensic psychiatric patients from other psychiatric patients [3].

Clinical pathways often begin in high‐security or forensic hospitals, where treatment should address both psychiatric needs and criminogenic risk factors [4]. Common diagnoses include schizophrenia, schizoaffective disorder, bipolar disorder, and personality disorders, frequently with comorbid substance use disorders [5, 6, 7, 8]. What further differentiates forensic psychiatric patients is their involvement in criminal or violent behaviors and the need to reduce the risk of reoffending [9]. Recent international research has highlighted the association between psychiatric disorders and violence, identifying risk factors and the role of risk assessment in forensic psychiatry [10].

General psychiatric patients, in contrast, engage with mental health services due to psychological suffering, functional impairment, or clinical risk, but without direct involvement in unlawful conduct [11]. Their treatment is typically provided in community mental health services or inpatient psychiatric units, and it is not subject to judicial obligations, treatment orders, or criminological risk assessments [12, 13]. Although the severity of psychiatric symptoms may be comparable to forensic psychiatric outpatients, they are not required to demonstrate reduced risk of reoffending. This highlights the need to evaluate outcomes beyond psychiatric symptoms, such as quality of life, general health, and functionality, to better address the long‐term needs of both groups.

Psychopathology, functionality, general health, and quality of life remain crucial domains for evaluating psychiatric results and treatment interventions. In general psychiatry, Quality of life (QoL) is recognized as an important indicator of recovery for both inpatients and outpatients. In forensic psychiatry, it is considered not only as an important outcome, but also as a predictor of criminal recidivism [14, 15]. Previous studies highlight the need for long‐term, individualized support for forensic psychiatric outpatients [16, 17].

Recent international studies have further advanced the understanding of forensic psychiatric outcomes. For example, Weber et al. [18] identified predictors of treatment response and risk mitigation in court‐mandated settings. Pav et al. [19] documented systemic and clinical factors impacting disparities in the length of stay across forensic inpatient units. A recent meta‐analysis by de Ruiter et al. [20] emphasized the role of shared assessment and management approaches while also noting the limited evidence linking these tools to violence outcomes. Complementing these international perspectives, national research from Greece has provided unique insights into the needs and pathways of forensic psychiatric populations [21, 22].

Nevertheless, longitudinal studies comparing forensic and general psychiatric patients remain scarce. To our knowledge this is the first prospective 24‐month study with four repeated assessment phases to examine longitudinal changes in quality of life, general health, functionality, and psychopathology in these two groups. The present study aims to address this gap by clarifying both the improvements achieved over time and the persistent disadvantages of forensic psychiatric outpatients. These findings are expected to inform tailored interventions and support the development of long‐term follow‐up care in forensic psychiatric services.

2. MATERIALS AND METHODS

2.1. Study design

A prospective longitudinal study was designed to assess the course of forensic psychiatric patients and general psychiatric patients over a 2‐year period. Data were collected over four phases: at baseline, and then at 6, 12, and 24 months. Forensic, social, psychopathological, and demographic data were collected, as well as repeated assessments of functionality, quality of life, general health, and psychopathology over time.

The prospective nature of the study ensures the systematic recording of the patients' course, allowing the identification of correlations between the parameters assessed as well as potentially influencing the future course of the patients. Long‐term follow‐up of psychiatric patients is critical for highlighting factors associated with relapse and recidivism [23]. Recording data over time allows more accurate detection of risk factors and contributes to improving the care of patients with psychiatric history.

2.2. Data collection

The study sample consisted of two groups: (a) forensic psychiatric outpatients and (b) general psychiatric patients with serious mental illness, monitored over a 24‐month period. All patients were followed at the outpatient forensic psychiatric unit of the 2nd Department of Psychiatry, University General Hospital “Attikon”.

Both forensic psychiatric outpatients and general psychiatric patients were living in the community. The general psychiatric group (psychiatric group) was the control group: it was composed of patients with similar diagnoses, without history of criminal or violent behavior. All patients were receiving antipsychotic treatment. Clinical characteristics including comorbid psychiatric diagnoses, were extracted from patients' medical records and confirmed by the treating psychiatrists, according to DSM‐5 diagnostic criteria. The basic characteristics of the two study groups were similar as shown in Table 1.

TABLE 1.

Demographic and clinical characteristics of forensic psychiatric outpatients and psychiatric patients.

	Total sample (n = 120; 100%)	Forensic psychiatric outpatients (n = 63; 52.5%)	Psychiatric patients (n = 57; 47.5%)	p
	n (%)	n (%)	n (%)
Gender
Male	99 (82.5)	54 (85.7)	45 (78.9)	0.330 a
Female	21 (17.5)	9 (14.3)	12 (21.1)
Ethnicity
Greek	113 (94.2)	58 (92.1)	55 (96.5)	0.443 b
Foreign	7 (5.8)	5 (7.9)	2 (3.5)
Relationship status
Married	15 (12.5)	7 (11.1)	8 (14)	0.596 b
Single	77 (64.2)	38 (60.3)	39 (68.4)
In a relationship	2 (1.7)	1 (1.6)	1 (1.8)
Separated	2 (1.7)	2 (3.2)	0 (0)
Divorced	24 (20)	15 (23.8)	9 (15.8)
Education
Elementary school	27 (22.5)	17 (27)	10 (17.5)	0.129 a
High school	33 (27.5)	20 (31.7)	13 (22.8)
Lyceum	46 (38.3)	22 (34.9)	24 (42.1)
Higher education	14 (11.7)	4 (6.3)	10 (17.5)
Current employment status
Employed	33 (27.7)	18 (28.6)	15 (26.8)	0.828 a
Unemployed	86 (72.3)	45 (71.4)	41 (73.2)
Residence
Urban area	112 (93.3)	59 (93.7)	53 (93)	0.847 b
Semi‐urban area	7 (5.8)	4 (6.3)	3 (5.3)
Rural area	1 (0.8)	0 (0)	1 (1.8)
Alone	33 (27.5)	21 (33.3)	12 (21.1)	0.136 b
Residence conditions
Family	34 (28.3)	15 (23.8)	19 (33.3)
Parental family	49 (40.8)	24 (38.1)	25 (43.9)
Homeless	3 (2.5)	3 (4.8)	0 (0)
Home care facilities	1 (0.8)	0 (0)	1 (1.8)
Past contact with mental health services
Yes	114 (95)	60 (95.2)	54 (94.7)	>0.999 b
No	6 (5)	3 (4.8)	3 (5.3)
Diagnosis
Schizophrenia – psychosis	118 (98.3)	63 (100)	55 (96.5)	0.224 b
Mood disorder	2 (1.7)	0 (0)	2 (3.5)
Comorbidity
Antisocial personality disorder	23 (19.2)	23 (36.5)	0 (0)	<0.001 b
Mixed personality disorder	5 (4.2)	5 (7.9)	0 (0)
Mental retardation	2 (1.7)	0 (0)	2 (3.5)
Schizotypal personality disorder	1 (0.8)	0 (0)	1 (1.8)
Substance use disorder	12 (10)	12 (19)	0 (0)
No comorbid diagnosis	77 (64.2)	23 (36.5)	54 (94.7)
History of Substance abuse
Yes	31 (25.8)	29 (46)	2 (3.5)	<0.001 b
No	89 (74.2)	34 (54)	55 (96.5)
	Mean (SD)	Mean (SD)	Mean (SD)
Age	41.3 (10.8)	42.8 (10.7)	39.8 (10.8)	0.126 c
Functionality scale (GAF) at baseline	51.0 (7.2)	50.38 (7.56)	51.61 (6.68)	0.348 c

^a Pearson's chi‐square test.

^b Fisher's exact test.

^c Student's t‐test.

Inclusion criteria: Forensic psychiatric outpatients were eligible to participate in the study if they had a diagnosis of serious mental illness, psychotic disorder, schizophrenia, schizoaffective disorder, mood disorder (according to DSM‐5 diagnostic criteria), and documented previous involvement with the criminal justice system. General psychiatric patients were eligible if they had a diagnosis of serious mental illness, psychotic disorder, schizophrenia, schizoaffective disorder, mood disorder, but no history of violent or aggressive behavior. All participants had to be over 18 years old, able to provide informed consent, and fluent in written and spoken Greek.

Exclusion criteria: Patients were excluded if their primary diagnosis was not a serious mental illness, psychotic disorder, schizophrenia, schizoaffective disorder, mood disorder (according to DSM‐5 diagnostic criteria), if they had significant cognitive impairments that could interfere with their ability to participate in the study, or if they lacked fluency in Greek.

A total number of 128 participants were initially included in the study. Eight individuals (n = 8; five forensic psychiatric patients and three general psychiatric patients) dropped out during follow‐up period and were excluded from the final analyses. Reasons for dropout were withdrawal of consent and relocation. Thus, the final sample consisted of 120 participants, all of whom completed the study with no missing data across the four assessment phases.

Data completeness was achieved through close clinical monitoring and integration of research assessments into the routine outpatient care, which minimized the fatigue and additional burden of them. The multidisciplinary team of the unit (psychiatrists, psychologists, social workers, and mental health nurses) ensured medication adherence, regular attendance at scheduled appointments, and provided support to both patients and their families. In parallel, the principal researcher coordinated all the assessments and maintained direct contact with the participants throughout the study. Missed appointments were followed up with phone call reminders. To ensure data completeness, participants were also asked to review their responses in real time and complete any missing items.

The integration of research procedure with therapeutic routine, combined with multidisciplinary support and therapeutic alliance, ensured continuous engagement and complete data over the 24‐month follow‐up period.

2.3. Treatment setting and clinical management

Participants in both groups were followed prospectively for 24 months within the outpatient service of a tertiary university hospital. The forensic outpatient group was monitored through the hospital's outpatient forensic psychiatric service, while the comparison group consisted of general psychiatric outpatients without a forensic history, receiving care within the same institutional framework.

Both groups were managed under a treatment‐as‐usual model, primarily based on antipsychotic pharmacotherapy, with comorbid psychiatric conditions addressed clinically on an individualized basis. No standardized or protocol‐based interventions specifically targeting comorbidity were implemented as part of the study. Management of comorbid conditions (e.g., substance use disorders or personality disorders) was conducted according to routine clinical judgment and standard outpatient practice in both groups. However, during regular one to one therapeutic sessions, comorbid disorders were actively monitored. In cases of symptomatic relapse or substance use, therapeutic individual intensity was increased, with particular emphasis placed on the potential legal consequences of such a relapse.

The study was designed to examine longitudinal outcomes in outpatient populations distinguished by forensic status rather than to compare predefined treatment protocols. Apart from forensic legal status, no systematic differences in therapeutic structure or intensity were introduced between groups. All participants were assessed at baseline, 6, 12, and 24 months using the same standardized instruments.

In routine practice, patients were scheduled for regular psychiatric follow‐up visits, with frequency determined by clinical needs. The multidisciplinary team included psychiatrists, psychologists, social workers, and mental health nurses, who collaborated in individualized care planning. Interventions included pharmacological management, psychoeducation, monitoring of treatment adherence, and coordination with family members and community services when required. No structured assertive outreach psychosocial intervention program was implemented as part of the study design; rather, care reflected standard outpatient clinical practice within the institutional setting.

2.4. Measurement

The psychometric measures that were used are:

1. World Health Organization Quality of Life—BREF version (WHOQOL‐BREF) is an psychometric assessment tool for quality of life, developed by the World Health Organization (WHO). It is a short version of the WHOQOL‐100 questionnaire and offers a broad assessment of quality of life. The Greek adaptation was validated by Ginieri—Coccossis et al. [24]. It assesses quality of life through four domains: Physical Health, Psychological Health, Social Relationships, Environment. It consists of 26 questions from the English original along with four other questions that resulted from the cultural adaptation of the tool in Greek data. It is completed using a five‐point Likert scale, which ranges from 1 (Very dissatisfied) to 5 (Very satisfied). Higher scores reflect an indication of better perceived quality of life by the participants [25].

2. General Health Questionnaire—28 (GHQ‐28) is a self‐administered questionnaire used to screen recent (last 2 weeks) psychiatric symptoms. The original version included 60 items, but a shorter version 28 items was developed [26, 27]. The present study used the 28‐item version, which includes four subscales. The subscales refer to somatic symptoms, anxiety and insomnia, social dysfunction and major depression. The GHQ‐28 is measured according to a Likert scale (0, 1, 2, 3). The Greek version of the GHQ‐28 was adapted to Greek cultural data and characteristics by Garyfallos et al. [28].

3. Global Assessment of Functioning (GAF) is a scale used by mental health professionals to assess an individual's global functioning. It provides a continuous rating scale ranging from 1 to 100, where 1 = severely impaired functioning and 100 = superior functioning. It is based on a hypothetical mental health‐illness continuum and assesses an individual's psychological, social, and occupational functioning. The GAF scale score contributes to the assessment and understanding of a patient's current state and the monitoring of changes over time, as it is guiding decisions about therapeutic support. The GAF Scale has been translated and validated in Greek psychiatric population by Madianos [29].

4. Positive and Negative Syndrome Scale (PANSS) is used to assess the psychopathology of patients and in particular the subscales of positive and negative symptoms, the scale composite, as well as general psychopathology [30]. PANSS has been translated and adapted in Greek [31]. It is frequently used in studies to assess the symptoms of patients with psychosis, and the collection of the information is carried out during the clinical interview. It includes 30 items, which consist of positive and negative symptoms of psychosis, as well as symptoms of general psychopathology, which are grouped into three categories. Scoring is conducted on a seven‐point scale (1 = not present, 7 = very severe).

5. A data collection questionnaire, which was created for the purposes of this study, recording demographic, forensic, social and clinical data, was also used.

2.5. Ethics

This study adhered to ethical standards; participants were informed of the research purpose, confidentiality, and anonymity. They were asked to give informed written consent to participate in the study. Participation in the study was entirely voluntary, and participants were informed that they could decline or withdraw their participation from the study at any time without any consequences to the continuation of their therapeutic support and care. Furthermore, participants were informed that the study data would be used for research purposes only.

The research study has received the necessary approval from the Research Ethics Committee of the Scientific Council of the University General Hospital “Attikon”.

2.6. Data analysis

Quantitative variables were expressed as mean values (Standard Deviation), while categorical variables were expressed as absolute and relative frequencies. For the comparison of proportions, chi‐square and Fisher's exact tests were used. Independent samples Student's t‐tests were used for the comparison of age between the two groups. Generalized linear models and especially repeated measures analysis of variance (ANOVA) were adopted to evaluate the changes observed in the scales under study among the two groups over the follow‐up period. Effect sizes via eta² were also computed from this analysis. Logarithmic transformations were used in the repeated measures analysis of variance (ANOVA). Bonferroni correction was used in case of multiple testing to control for type I error. All reported p values are two‐tailed. Statistical significance was set at p < 0.05. Analyses were conducted using SPSS statistical software (version 27.0).

3. RESULTS

Data from 120 patients were collected and analyzed (63 forensic psychiatric outpatients and 57 general psychiatric patients), predominantly male (82.7%), with a mean age of 41.3 years (SD = 10.8 years). As shown in Table 1, the two groups had similar demographics. Significant differences were found between comorbid diagnoses (p < 0.001). Psychiatric patients showed less comorbidity, while forensic psychiatric outpatients reported significantly higher rates of substance use disorders.

As shown in Table 2, most forensic psychiatric outpatients had a history of imprisonment, between 1 and 10 years. Over half of them had committed a crime against a known victim, and almost half of them reported a prior criminal history before the index offense.

TABLE 2.

Forensic and criminological characteristics of forensic psychiatric outpatients.

	n (%)
Imprisonment
Yes	45 (93.8)
No	3 (6.3)
Number of victims
0	5 (10.4)
1	32 (66.7)
2	7 (14.6)
3	4 (8.3)
Relationship between offender and victim
Spouses	3 (7)
Relatives	2 (4.7)
Relationship	3 (7)
Professional	1 (2.3)
Simple contact	22 (51.2)
Unknown	12 (27.9)
Prior criminal history
Yes	21 (43.8)
No	27 (56.3)

As shown in Table 3, no significant differences were found between the two groups in the GAF scale (p > 0.05) and in all subscales of WHOQOL‐BREF, except for the “Environment” domain, which was significantly higher in psychiatric patients at baseline and 6 months. After applying the Bonferroni correction, both groups showed significant improvements over time, with steadily increasing scores at each measurement. The overall degree of improvement was similar in the two groups (p > 0.05), although the increase in the “Environment” domain was significantly greater among the forensic psychiatric outpatients (Figure 1).

TABLE 3.

Longitudinal changes in GAF and WHOQoL‐BREF scales among forensic psychiatric outpatients and psychiatric patients over a 24‐month follow‐up.

	Baseline T1	6 months T2	12 months T3	24 months T4	Change from Τ1 to Τ4	p (eta2) b	p (eta2) c
	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
Functionality scale (GAF)
Forensic psychiatric patient	50.38 (7.56)t2.t3.t4	55.38 (6.89)t1.t3.t4	59.17 (7.02)t1.t2.t4	62.3 (7.69)t1.t2.t3	11.92 (6.31)	<0.001 (0.665)	0.695 (0.002)
Psychiatric patient	51.61 (6.68)t2.t3.t4	56.51 (6.26)t1.t3.t4	60.54 (6.83)t1.t2.t4	64.33 (7.38)t1.t2.t3	12.72 (6.34)	<0.001 (0.659)
p (eta 2 ) a	0.304 (0.009)	0.329 (0.008)	0.284 (0.010)	0.143 (0.018)
Overall QoL/general health
Forensic psychiatric patient	33.53 (16.92)t2.t3.t4	46.23 (15.98)t1.t3.t4	58.33 (12.5)t1.t2.t4	68.85 (10.26)t1.t2.t3	35.32 (15.64)	<0.001 (0.561)	0.747 (0.001)
Psychiatric patient	34.87 (14.7)t2.t3.t4	45.18 (15.08)t1.t3.t4	58.33 (12.35)t1.t2.t4	69.08 (10.33)t1.t2.t3	34.21 (13.88)	<0.001 (0.553)
p (eta 2 ) a	0.742 (0.001)	0.850 (0.000)	0.975 (0.000)	0.927 (0.000)
Physical health
Forensic psychiatric patient	45.11 (14.84)t2.t3.t4	53.09 (14.01)t1.t3.t4	60.32 (9.47)t1.t2.t4	65.52 (6.87)t1.t2.t3	20.41 (10.76)	<0.001 (0.474)	0.504 (0.005)
Psychiatric patient	43.27 (14.66)t2.t3.t4	53.12 (13.26)t1.t3.t4	60.09 (10.61)t1.t2.t4	65.4 (7.57)t1.t2.t3	22.12 (10.35)	<0.001 (0.505)
p (eta 2 ) a	0.502 (0.004)	0.858 (0.000)	0.766 (0.001)	0.877 (0.000)
Mental health
Forensic psychiatric patient	39.95 (14.75)t2.t3.t4	48.48 (13.22)t1.t3.t4	59.13 (10.68)t1.t2.t4	65.41 (9.28)t1.t2.t3	25.46 (12.21)	<0.001 (0.563)	0.777 (0.002)
Psychiatric patient	38.38 (14.96)t2.t3.t4	49.56 (15.68)t1.t3.t4	58.48 (12.77)t1.t2.t4	64.77 (10.65)t1.t2.t3	26.39 (10.54)	<0.001 (0.549)
p (eta 2 ) a	0.592 (0.002)	0.946 (0.000)	0.566 (0.003)	0.628 (0.002)
Social relationships
Forensic psychiatric patient	29.44 (15.81)t2.t3.t4	40.87 (15.31)t1.t3.t4	49.44 (11.71)t1.t2.t4	55.63 (11.55)t1.t2.t3	26.19 (13.55)	<0.001 (0.460)	0.680 (0.002)
Psychiatric patient	32.11 (15.03)t2.t3.t4	41.58 (15.9)t1.t3.t4	51.4 (13.35)t1.t2.t4	54.82 (11.88)t1.t2.t3	22.72 (9.96)	<0.001 (0.374)
p (eta 2 ) a	0.809 (0.000)	0.754 (0.001)	0.652 (0.002)	0.606 (0.002)
Environment
Forensic psychiatric patient	44.59 (14.73)t2.t3.t4	52.28 (10.84)t1.t3.t4	60.81 (8.73)t1.t2.t4	65.13 (7.96)t1.t2.t3	20.54 (12.29)	<0.001 (0.551)	0.023 (0.036)
Psychiatric patient	50.49 (12.33)t2.t3.t4	58.44 (10.07)t1.t3.t4	64.36 (9.97)t1.t2.t4	67.38 (7.56)t1.t2.t3	16.89 (9.11)	<0.001 (0.333)
p (eta 2 ) a	0.016 (0.048)	0.005 (0.066)	0.110 (0.022)	0.132 (0.019)

Note: Analyses were made with scores' logarithmical transformations. t1, t2, t3, t4: significant differences with T1, T2, T3 and T4 timepoints, respectively, after Bonferroni correction. Bold values indicate statistical significance (p < 0.05).

^a p‐value and effect size (eta²) for group comparisons.

^b p‐value and effect size (eta²) for time comparisons.

^c Repeated measures ANOVA p‐value and effect size (eta²), regarding time × group effect.

FIGURE 1.

Longitudinal changes in “Environment” domain of the WHOQOL‐BREF among forensic psychiatric outpatients and psychiatric patients over 24‐month follow‐up.

As shown in Table 4, no significant differences were found between the two groups in the “Social Symptoms,” “Social Dysfunction” and “Severe Depression” scores (p > 0.05). In contrast, psychiatric patients scored significantly higher in “Anxiety/Insomnia” subscale across all measurements. Additionally, the total GHQ‐28 score was significantly greater in this group at 6 and 24 months. When comparisons were made over time within each group, both groups showed decreases in scores. After applying the Bonferroni correction, significant differences were observed between all measurement points, indicating that scores declined over the duration of the study. As indicated by interaction effects, both groups showed similar reductions in “Social Symptoms” (p = 0.130), but for other dimensions and the total score (Figure 2), psychiatric patients had significantly greater decreases than forensic psychiatric outpatients during the follow‐up.

TABLE 4.

Longitudinal changes in GHQ‐28 scale among forensic psychiatric outpatients and psychiatric patients over a 24‐month follow‐up.

	Baseline T1	6 months T2	12 months T3	24 months T4	Change from T1 to T4	p (eta2) b	p (eta2) c
	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
Somatic symptoms
Forensic psychiatric patient	6.48 (3.96)t2.t3.t4	4.1 (2.97)t1.t3.t4	3.19 (2.01)t1.t2.t4	2.62 (1.34)t1.t2.t3	−3.86 (3.21)	<0.001 (0.696)	0.130 (0.017)
Psychiatric patient	7.11 (4.85)t2.t3.t4	3.95 (3.11)t1.t3.t4	3.21 (2.54)t1.t2.t4	2.56 (2.08)t1.t2.t3	−4.54 (3.32)	<0.001 (0.742)
p (eta 2 ) a	0.737 (0.001)	0.613 (0.002)	0.618 (0.002)	0.289 (0.010)
Anxiety/insomnia
Forensic psychiatric patient	10.56 (3.31)t2.t3.t4	7.29 (2.99)t1.t3.t4	5.11 (2.46)t1.t2.t4	4.38 (1.85)t1.t2.t3	−6.17 (2.45)	<0.001 (0.748)	<0.001 (0.063)
Psychiatric patient	9.4 (4.14)t2.t3.t4	5.05 (3.22)t1.t3.t4	3.63 (2.76)t1.t2.t4	2.96 (2.12)t1.t2.t3	−6.44 (3.26)	<0.001 (0.816)
p (eta 2 ) a	0.035 (0.037)	<0.001 (0.140)	<0.001 (0.101)	<0.001 (0.137)
Social dysfunction
Forensic psychiatric patient	14.6 (3.62)t2.t3.t4	11.54 (3.34)t1.t3.t4	8.75 (2.18)t1.t2.t4	7.46 (1.31)t1.t2.t3	−7.14 (3.17)	<0.001 (0.840)	0.002 (0.044)
Psychiatric patient	15.47 (3.15)t2.t3.t4	10.67 (3.01)t1.t3.t4	8.79 (2.14)t1.t2.t4	7.53 (1.31)t1.t2.t3	−7.95 (2.8)	<0.001 (0.863)
p (eta 2 ) a	0.112 (0.021)	0.150 (0.018)	0.840 (0.000)	0.728 (0.001)
Severe depression
Forensic psychiatric patient	2.46 (2.94)t2.t3.t4	1.19 (2.31)t1.t3.t4	0.79 (1.71)t1.t2.t4	0.38 (0.77)t1.t2.t3	−2.08 (2.6)	<0.001 (0.360)	0.046 (0.026)
Psychiatric patient	3.53 (3.63)t2.t3.t4	1.4 (2.4)t1.t3.t4	0.95 (1.76)t1.t2.t4	0.58 (1.24)t1.t2.t3	−2.95 (3)	<0.001 (0.487)
p (eta 2 ) a	0.056 (0.031)	0.623 (0.002)	0.628 (0.002)	0.481 (0.004)
Total score GHQ‐28
Forensic psychiatric patient	34.1 (11.52)t2.t3.t4	24.11 (9.16)t1.t3.t4	17.84 (6.37)t1.t2.t4	14.84 (3.82)t1.t2.t3	−19.25 (9.44)	<0.001 (0.862)	<0.001 (0.071)
Psychiatric patient	35.51 (14.45)t2.t3.t4	21.07 (10.05)t1.t3.t4	16.58 (7.95)t1.t2.t4	13.63 (5.82)t1.t2.t3	−21.88 (10.78)	<0.001 (0.894)
p (eta 2 ) a	0.739 (0.001)	0.028 (0.040)	0.107 (0.022)	0.030 (0.039)

Note: Analyses were made with scores' logarithmical transformations. t1, t2, t3, t4: significant differences with T1, T2, T3 and T4 timepoints, respectively, after Bonferroni correct. Bold values indicate statistical significance (p < 0.05).

^a p‐value and effect size (eta²) for group comparisons.

^b p‐value and effect size (eta²) for time comparisons.

^c Repeated measures ANOVA p‐value and effect size (eta²), regarding time × group effect.

FIGURE 2.

Longitudinal changes in total GHQ‐28 score among forensic psychiatric outpatients and psychiatric patients over a 24‐month follow‐up.

Changes in PANSS scales, by group are presented in Table 5. No significant differences were observed between the groups in the Negative scale, General psychopathology, or Total PANSS score (p > 0.05). However, forensic psychiatric outpatients consistently had higher Composite scores, and their Positive scale scores were significantly higher at 12 and 24 months. Within group analyses showed significant decreases in all PANSS scores for both groups, except for the Composite scale. After Bonferroni correction, both groups showed significant reductions across all PANSS scales. The overall degree of change in PANSS scales during the follow‐up was similar in the two groups as indicated from the nonsignificant interaction effects.

TABLE 5.

Longitudinal changes in PANSS scale among forensic psychiatric outpatients and psychiatric patients over 24‐month follow‐up.

	Baseline T1	6 months T2	12 months T3	24 months T4	Change from T1 to T4	p (eta2) b	p (eta2) c
	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
Positive scale
Forensic psychiatric patient	20.46 (7.73)t2.t3.t4	16.29 (7.22)t1.t3.t4	13.73 (6.29)t1.t2.t4	11.87 (5.31)t1.t2.t3	−8.59 (4.45)	<0.001 (0.626)	0.075 (0.023)
Psychiatric patient	18.82 (7.55)t2.t3.t4	14.23 (6.81)t1.t3.t4	11.44 (5.21)t1.t2.t4	9.86 (4.4)t1.t2.t3	−8.96 (4.76)	<0.001 (0.680)
p (eta 2 ) a	0.469 (0.004)	0.067 (0.028)	0.015 (0.049)	0.006 (0.061)
Negative scale
Forensic psychiatric patient	20.49 (7.49)t2.t3.t4	16.52 (6.73)t1.t3.t4	13.9 (5.82)t1.t2.t4	12.17 (5.18)t1.t2.t3	−8.32 (4.07)	<0.001 (0.746)	0.264 (0.011)
Psychiatric patient	21.04 (7.35)t2.t3.t4	16.28 (6.4)t1.t3.t4	13.6 (5.65)t1.t2.t4	12.07 (4.35)t1.t2.t3	−8.96 (4.08)	<0.001 (0.731)
p (eta 2 ) a	0.498 (0.004)	0.853 (0.000)	0.799 (0.001)	0.884 (0.000)
Composite scale
Forensic psychiatric patient	−0.03 (3.78)	−0.24 (3.71)	−0.17 (3.39)	−0.3 (2.61)	−0.27 (2.77)	0.936 (0.004)	0.635 (0.004)
Psychiatric patient	−2.21 (4.06)	−2.05 (4.49)	−2.16 (3.21)	−2.21 (2.81)	0 (3.31)	0.593 (0.016)
p (eta 2 ) a	0.004 (0.070)	0.007 (0.059)	0.004 (0.069)	<0.001 (0.112)
General psychopathology
Forensic psychiatric patient	35.27 (10.89)t2.t3.t4	28.63 (10.3)t1.t3.t4	24.78 (8.01)t1.t2.t4	22.27 (7.42)t1.t2.t3	−13 (6.26)	<0.001 (0.843)	0.143 (0.016)
Psychiatric patient	36.98 (14.28)t2.t3.t4	28.53 (10.41)t1.t3.t4	24.33 (8.54)t1.t2.t4	22.12 (6.2)t1.t2.t3	−14.86 (9.12)	<0.001 (0.855)
p (eta 2 ) a	0.616 (0.002)	0.865 (0.000)	0.624 (0.002)	0.999 (0.000)
Total PANSS score
Forensic psychiatric patient	75.71 (23.78)t2.t3.t4	61.44 (22.75)t1.t3.t4	51.79 (19.4)t1.t2.t4	45.86 (15.77)t1.t2.t3	−29.86 (11.96)	<0.001 (0.855)	0.115 (0.018)
Psychiatric patient	76.84 (27.95)t2.t3.t4	59.04 (22.27)t1.t3.t4	49.37 (18.46)t1.t2.t4	44.05 (14.09)t1.t2.t3	−32.79 (16.58)	<0.001 (0.861)
p (eta 2 ) a	0.932 (0.000)	0.462 (0.005)	0.432 (0.005)	0.478 (0.004)

Note: Analyses were made with scores' logarithmical transformations. t1, t2, t3, t4: significant differences with T1, T2, T3 and T4 timepoints, respectively, after Bonferroni correction. Bold values indicate statistical significance (p < 0.05).

^a p‐value and effect size (eta²) for group comparisons.

^b p‐value and effect size (eta²) for time comparisons.

^c Repeated measures ANOVA p‐value and effect size (eta²), regarding time × group effect.

4. DISCUSSION

The present study assessed the long‐term outcomes of forensic psychiatric outpatients compared with general psychiatric patients across four domains: quality of life, psychopathology, functionality, and general health. Both groups showed significant improvement across all domains over the 24‐month follow‐up, underlining the importance of consistent and stable outpatient psychiatric care.

Forensic psychiatric outpatients consistently reported lower outcomes, which reflect their greater baseline disadvantages in these domains. To our knowledge, this is a first 2‐year longitudinal study with four repeated assessments comparing these two groups, which strengthens the novelty and contribution of the findings.

Both groups reported limited support networks, though this was more prominent in forensic psychiatric outpatients. Schizophrenia was the primary diagnosis in both groups, while forensic psychiatric outpatients had higher rates of comorbid antisocial personality disorder, mixed personality disorder, and substance use disorder. These findings are consistent with previous research linking these factors to increased risk of violent or criminal behavior [32, 33, 34].

High unemployment, low educational attainment, and social isolation were also present in both groups. These findings are in line with earlier findings on the role of socioeconomic disadvantage and low educational attainment in the development of mental disorders, crime, relapse, recidivism, and barriers to early intervention and mental health services [35, 36, 37, 38, 39, 40, 41, 42, 43]. Living with family may represent a protective factor, although the quality of family relationships remains critical [44]. Conversely, employment is associated with better quality of life and self‐esteem [45, 46], highlighting the importance of rehabilitation and social integration as key factors in long‐term outpatient care.

4.1. Quality of life and functioning

Both groups showed improvement in QoL and functioning across all domains over the 24‐month follow‐up, confirming the positive impact of the long‐term outpatient care. Forensic psychiatric outpatients demonstrated significant improvement in the “environment” dimension. These findings are consistent with previous evidence, emphasizing the importance of therapeutic environments, safe housing, positive staff relationships, and prosocial peer connections, which enhance autonomy, agency, and emotional well‐being [47, 48].

Interventions including volunteer‐led social support into standard forensic care have also shown to improve mental well‐being, QoL, social participation, and self‐sufficiency [49]. At the same time, prior studies have shown that low QoL is associated with major mental disorders [50, 51], unmet needs [52, 53], and consequently lower levels of global functioning [54, 55]. Forensic psychiatric outpatients started with lower baseline scores, which may indicate greater baseline disadvantages, but also greater potential for improvement. This suggests that structured outpatient care can enhance living conditions and daily functioning.

The longitudinal changes observed in the present study are broadly consistent with findings reported in the literature on forensic outpatient psychiatric care. Previous studies have demonstrated that forensic psychiatric outpatients, despite presenting with higher baseline clinical complexity and comorbidity, may exhibit gradual improvements in symptom severity, global functioning, and quality of life when sustained treatment is maintained [10, 14, 18].

Similarly, longitudinal outpatient studies in severe mental illness populations indicate that structured, multidisciplinary follow‐up within routine clinical practice is associated with progress in stabilization over extended follow‐up periods [56, 57, 58]. In line with these findings, both forensic and general psychiatric outpatients in our sample showed significant improvements across psychopathology, functioning, and general health measures over 24 months, although the forensic psychiatric group consistently demonstrated greater overall clinical burden.

The persistence of between‐group differences despite longitudinal improvement supports prior evidence suggesting that treatment‐as‐usual outpatient care may promote the stabilization of the symptomatology, but may be insufficient to fully address the multidimensional needs of forensic psychiatric populations [20]. Taken together, our findings extend existing literature by providing prospective data from a Greek forensic outpatient setting and reinforce the importance of structured, sustained community‐based follow‐up.

4.2. Psychopathology and symptom reduction

In this study, both groups showed significant improvement in psychopathology over the 24‐month follow‐up, but general psychiatric patients demonstrated greater improvement in positive symptoms on the PANSS positive scale. This difference may reflect illness related aspects, such as chronicity, complexity, and other psychosocial factors including stigma, social isolation, and limited social support. The involvement of forensic psychiatric patients with the criminal justice system may further complicate recovery pathways in this group. These findings are consistent with previous studies showing that improvements in social functioning and functional remission, premorbid personality adjustment, social networks, and shorter duration of untreated psychosis are associated with better long‐term outcomes for patients with schizophrenia [56, 57, 58, 59, 60, 61].

No significant differences were observed between the two groups in social symptoms, social dysfunction, or severe depression. Psychiatric patients reported higher anxiety/insomnia and total GHQ score was significantly higher at 6 and 24 months, which may reflect different symptom patterns between the two groups. Sleep disturbances remain a persistent difficulty; previous studies report high rates of insomnia in forensic psychiatric patients (20%–30%) [62, 63, 64]. Considering that sleep problems are associated with the development, severity, and relapse of mental health disorders [65], their systematic monitoring should be included in the long‐term outpatient care.

Although forensic psychiatric outpatients improved steadily across all domains, their improvement in some domains was slower. Factors such as comorbidity, social difficulties, and prior involvement with the criminal justice system may contribute to these differences. In addition, unmet physical, mental, and social needs [66, 67, 68], which may further limit their recovery. Studies suggest that psychosocial contacts with higher levels of respect, dignity, staff relationships, and family contact are associated with lower psychological distress and better general mental health [69, 70, 71].

Overall, these results indicate that outpatient long‐term care contributes to psychopathology improvement in both groups, though forensic psychiatric outpatients may remain more vulnerable due to the impact of clinical and social factors. These findings highlight the importance of integrating psychosocial and rehabilitative interventions alongside pharmacological treatment in long‐term forensic outpatient care.

4.3. Clinical implications

Our study showed that both forensic psychiatric outpatients and general psychiatric patients improved during the 24‐month long‐term follow‐up. These results highlight the importance of long‐term outpatient psychiatric care that addresses not only the reduction of psychiatric symptoms and medication adherence but also the daily life challenges.

Nevertheless, forensic psychiatric outpatients continue to face greater disadvantages in quality of life, general health, functionality, and psychopathology. These difficulties reflect the impact of comorbidity, limited support network, as well as socioeconomic problems. Stigma and legal problems add further barriers [72, 73], reducing autonomy and discouraging employment or social engagement. These are known risk factors for relapse and recidivism and may explain the less improvement observed in this group. Including social network support into forensic outpatient care has been shown to improve quality of life, social interaction, well‐being, and may contribute to reducing recidivism [49].

In our setting, the multidisciplinary team played a central role in ensuring patients' engagement, medication adherence, and regular attendance on preagreed appointments. Close collaboration among psychiatrists, psychologists, social workers, and mental health nurses enhanced continuous monitoring, multilevel assessment, and individualized therapeutic planning. Consistent follow‐up, psychoeducation for individuals and families, active encouragement to pursue employment and/or education, and coordination of physical health care were standard practices. Applied systematically for more than 15 years in this unit, these practices helped to build trust and supported long‐term adherence.

Overall, our findings highlight the need for community‐based forensic psychiatric services that integrate individualized psychiatric care, vocational rehabilitation, and family and social support. Innovative tools, such as telepsychiatry, may further improve access, support, and treatment adherence [74]. Multidisciplinary, structured services could mitigate disadvantages, strengthen reintegration into the community, and reduce the risk of relapse and reoffending.

4.4. Strengths, limitations and future directions

This study had several strengths. The prospective longitudinal design, in four distinct phases over a 24‐month period, allowed the assessment of changes and improvements in different domains. This provided the possibility to monitor patients' progress in long‐term community care. Data was collected in a community setting and not in a restricted environment. This increases the validity of the findings, as the results refer to real life conditions and strengthens the applicability of the findings to clinical practice. Including a comparison group of general psychiatric patients increased internal validity, clarifying risk factors and differences between the two groups. Finally, the use of validated, multidimensional tools ensured reliable assessment of both clinical and functional outcomes.

The longitudinal design also reduces potential evaluative distortions (such as time or gender bias), often present in forensic psychiatric care [75]. Repeated assessments over time contributed to reducing evaluation distortions about patients' improvements and informed both clinical and legal decisions. The combination of a prospective design and repeated assessments represents a clear methodological strength.

This study has also several limitations. Although organized forensic outpatient services in Greece are limited and primarily concentrated in major urban centers, the single‐site design may limit national representativeness, particularly for regions without structured forensic outpatient follow‐up. Furthermore, the predominantly Greek sample may limit generalizability to countries with different healthcare and forensic legal systems. Differences in service organization, legal supervision frameworks, and community‐based mental health structures may influence longitudinal outcomes.

Comorbidity was more prevalent in the forensic outpatient group, yet comorbidity‐specific interventions were not systematically recorded, and residual confounding cannot be excluded. Both groups were managed under a treatment‐as‐usual approach within the same institution; however, contextual differences inherent to forensic status (e.g., legal oversight or supervision requirements) may have influenced outcomes but were not systematically reordered as measurable variables. Furthermore, the absence of an external routine‐care comparison cohort limits direct cross‐study comparisons.

The sample size of the forensic psychiatric outpatients was sufficient for statistical analyses but may limit the assessment of subgroup differences and interaction effects. Participants were not randomly selected, which may further restrict generalizability. The use of self‐report measures may have introduced response bias, including social desirability, prejudices, or therapeutic influence, particularly in quality of life and perceived well‐being. While the 24‐month follow‐up period was adequate to detect significant improvements, longer observation would be necessary to assess outcomes such as recidivism, long‐term medication adherence, and sustained community reintegration.

Future research should include larger, randomized, multisite, and potentially randomized samples to strengthen external validity. A longer follow‐up period would also contribute to better monitor the stability of the outcomes. Incorporation of qualitative methodologies, including perspectives of patients, caregivers, and mental health professionals, may contribute to more comprehensive and individualized models of forensic outpatient care.

5. CONCLUSIONS

Forensic psychiatric patients with psychotic disorders require long‐term support and care. Community‐based treatment can enhance quality of life, functioning, general health, and psychopathological outcomes. Improvements in these domains are associated with reductions in aggression, impulsivity, and risk of relapse and recidivism.

The findings of this study emphasize the importance of delivering therapeutic interventions in community settings. Care models should integrate quality of life, functionality, and relapse prevention as core objectives in the care of these individuals. Early intervention combined with continuous monitoring is essential for long‐term clinical stability and successful reintegration into society.

Overall, this study shows that long‐term outpatient care benefits both forensic psychiatric outpatients and general psychiatric patients, but forensic psychiatric outpatients face disadvantages due to social and clinical factors. Community‐based interventions that include psychiatric and social support, as well as rehabilitation, are necessary to promote reintegration.

These findings provide important insights for designing targeted interventions and contribute to the development of evidence‐based policies and best practices in forensic psychiatric care.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

The study was conducted in accordance with the Declaration of Helsinki and approved by the Bioethical and Deontological Committee of General University Hospital “Attikon,” National and Kapodistrian University of Athens—approval code EB∆502/17‐7‐18, on July 17, 2018.

INFORMED CONSENT STATEMENT

Informed consent was obtained from all subjects involved in the study.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.